UnisonPersistent Memory for LlamaIndex - Unison Integration
Drop-in agent memory for LlamaIndex: UnisonMemory replaces ChatMemoryBuffer and recalls team knowledge before each turn, ingesting what the agent learns.
Learn how to give LlamaIndex agents persistent team memory with Unison.
Official adapter
The fastest path is the llama-index-memory-unison package — a BaseMemory implementation that drops in wherever you'd use ChatMemoryBuffer:
pip install llama-index-memory-unisonfrom llama_index.memory.unison import UnisonMemory
from llama_index.core.agent import ReActAgent
from llama_index.llms.openai import OpenAI
memory = UnisonMemory.from_client(session_id="user-123", api_key="usk_live_...")
agent = ReActAgent.from_tools(tools, llm=OpenAI(), memory=memory)On every turn it recalls relevant context from your brain (GET /v1/brain/context) and injects it as a system message, then ingests new turns (POST /v1/brain/ingest) so entity resolution and bitemporal facts accumulate. UNISON_TOKEN / UNISON_API_URL are read from the environment if not passed.
Replacing ChatMemoryBuffer
# before
from llama_index.core.memory import ChatMemoryBuffer
memory = ChatMemoryBuffer.from_defaults(token_limit=3000)
# after — same slot, now shared and persistent across sessions and teammates
from llama_index.memory.unison import UnisonMemory
memory = UnisonMemory.from_client(session_id="user-123", api_key="usk_live_...")Unlike a conversation buffer, what lands in the brain is shared: every teammate's agent recalls it. See the package README for the full API.
Persistent Memory for LangChain - Unison Integration
Add persistent team memory to LangChain agents and chains: recall before invocation, ingest after - shared across every session and teammate.
Shared Memory for CrewAI - Unison Integration
Give a CrewAI crew persistent shared memory: every agent reads the same project facts, and nothing is re-learned between runs.